Test Patterns

Forum for discussion of narrow-bandwidth mechanical television

Moderators: Steve Anderson, Dave Moll, Andrew Davie

Test Patterns

Postby Andrew Davie » Tue May 15, 2007 10:58 pm

Here's my attempt at a test pattern.

The idea behind this one is to include a recognisable image -- I chose Baird over Nipkow because he is much more associated with actual television (in my mind, at least).

The test pattern uses a face to give some idea of "real world" usage. It also includes a greyscale range at the bottom, and both horizontal and vertical 'resolution' indicators. These are basically alternating black/white dots at what I consider optimal resolution (32 x 48 pixels, with a 1:1 pixel aspect ratio).

Finally, the image includes horizontal lines in black and white, and vertical lines in black and white. I can't at this stage think of much more that would be required for a test pattern. Please feel free to modify/improve.
Attachments
testpatternbaird.png
testpatternbaird.png (2.23 KiB) Viewed 7231 times
testpatternbairdbig.png
testpatternbairdbig.png (3.62 KiB) Viewed 7231 times
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1512
Joined: Wed Jan 24, 2007 4:42 pm
Location: Hobart, Tasmania, Australia

Postby Klaas Robers » Wed May 16, 2007 12:58 am

Andrew, if you make the pattern in 256 x 412, eventually in colour, to start with in .bmp, then I think that I can convert it into CCNC video standard. My program makes then a .wav file of a length of 2 seconds, which you can copy to any length. As you might have read this wav file can be played also in black and white.

All low pass filtering is done by the program, it makes of it what is possible. And then, if I remember right, it can also make a B+W picture back, but now limited to 32 lines horizontally and 10 kHz vertically.

I will add an example, so you have the correct outline. It is in .gif, because .bmp isn't accepted by the forum. I convert them always back to .bmp, that is the input format for my program.
Attachments
grant-pic.gif
Grant Dixon testpattern with colour bars
grant-pic.gif (6.75 KiB) Viewed 7222 times
grant-bw.gif
This is the file converted back to black and white and in 32 lines.
grant-bw.gif (11.63 KiB) Viewed 7220 times
User avatar
Klaas Robers
Frankenstein was my uncle.
 
Posts: 1498
Joined: Wed Jan 24, 2007 8:42 pm
Location: Valkenswaard, the Netherlands

Postby DrZarkov » Wed May 16, 2007 1:10 am

With GraphicConverter on Macintosh I can convert anything viewable into a Quicktime movie of theoreticly any length. This can be converted to a NBTV video. I will try it when I'm back home and my wife lets me use the Mac (mostly I have to use an old Dell notebook...).
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1037
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Postby Andrew Davie » Wed May 16, 2007 1:41 am

Klaas Robers wrote:Andrew, if you make the pattern in 256 x 412, eventually in colour, to start with in .bmp, then I think that I can convert it into CCNC video standard. My program makes then a .wav file of a length of 2 seconds, which you can copy to any length. As you might have read this wav file can be played also in black and white.


Thanks for the above. That's very clever how the colour bars view as a greyscale in the black and white version.

When there are file formats that are not allowed in the forum, you can always zip up the files, and post the zip.

Last I knew, .bmp was not a 'web' format -- it is not supported by browsers on other platforms (or, it wasn't -- it's been a while since I played with web stuff). Basically, it used to be a "forbidden" format to use, so I didn't specifically enable it. I try to steer clear of GIF files as a matter of principle, because of the patent issues involved with the compression scheme. In any case, PNG is a far superior format.

As to conversion... I had assumed that I could make the original test pattern in *any* size (but I have kept it to 32 x 48 just for ease of drawing), and then add the frame to a track using some video editor, save that out as a movie as long as I wish, then run that movie (avi/wmv, etc) through Video2NBTV -- and I'd have the CCNC version I needed.

I might try this first, and will report on the results.
User avatar
Andrew Davie
"Gomez!", "Oh Morticia."
 
Posts: 1512
Joined: Wed Jan 24, 2007 4:42 pm
Location: Hobart, Tasmania, Australia

Postby DrZarkov » Wed May 16, 2007 5:28 am

Taken from the other thread because it's fitting here much better: My wife photoshoped for me the BBC testcard with photos of Baird and Nipkow. They are looking great in normal resolution, but for 32 lines I have a lot of work on my Amiga using a pixel-oriented painting program (PPaint). I tried simply to scale it down on the PC, but it's not only looking awful, it's useless.
Attachments
bairdtest.gif
bairdtest.gif (41.71 KiB) Viewed 7204 times
nipkowtest-02.gif
nipkowtest-02.gif (45.3 KiB) Viewed 7204 times
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1037
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Postby DrZarkov » Thu May 17, 2007 5:35 am

My wife saw me drawing, and she said that our testcards are looking so "eighties", we should take something more old-fashioned. And she said that a moving picture would make a better testcard. She made one example with Photoshop as a gif-anim. I was not able to transfer it into anything which "AVI to NBTV" will convert until now, maybe somebody knows a program for that. Most video converting programs have a big problem with a video of only 32 x 48 pixels...
Attachments
testbild-anim.gif
testbild-anim.gif (67.61 KiB) Viewed 7187 times
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1037
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

NBTV test strip?

Postby Stephen » Thu May 17, 2007 5:45 am

Your wife has an interesting idea, Volker. To keep it simple and mechanical, perhaps the best thing to do would be to make a NBTV test "strip" instead of a card and wrap it around a slowly revolving drum in front of the camera. The strip could then contain much more test pattern information than would be possible in the limited area of a 30/32 line test card.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Postby DrZarkov » Thu May 17, 2007 7:27 am

The idea with the drum is good, like that what the BBC used in that time, as to be seen on that testcard-site http://www.pembers.freeserve.co.uk/Test-Cards/index.html#30-Lines

(and not unlike much more modern testcards and logos. Many graphical tricks at the BBC and other TV stations which looked very modern and computer-made, like the BBC globe in the early 70th were in fact just models made of paper and plastic, animated by a motor. The earliest trick like that was I think the testcard with the clock.)

But until now I don't have a NBTV camera (this will follow later this year), so I'm happy that I have access to modern technology like computers and cd-rom.

I've found a picture of the first german TV test movie, a 10 minute silent movie calles "Wochenende" (weekend) with Imogen Orkutt and Schura von Finkelstein. They are singing (in a silent movie!?) "Horch, was kommt von draußen rein". This film had been used from 1929 to around 1934.
Attachments
29-test.gif
29-test.gif (18.77 KiB) Viewed 7175 times
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1037
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Test cards and clocks.

Postby Steve Anderson » Thu May 17, 2007 1:26 pm

The earliest trick like that was I think the testcard with the clock.


..or just the clock on its own. This was used as 'filler' to make up some seconds that might have been gained or lost in the previous programme. This ensured that The Six O'clock news started exactly on time. Apart from the time-pips on radio, it was one of the few ways the public could accurately set their clocks. There was the GPO 'speaking clock', but that cost money to make the call (I think it still does).

Nowadays real time is rarely seen on TV. The local terrestrial services do here in the morning, but on the satellite service it's omitted. This is due to the delay in the signal going up to the 'bird' and back down again.

On international circuits there's not only that delay (especially if involves more than one 'hop'), but those caused by frame synchroinsers and standards convertors. There is also different time zones to consider.

Advertisers didn't like the clock, it reminded people that they were late for work and therefore turned off their TV.

Nowadays we have GPS and a few other accurate references for time like MSF. Even the Internet is good enough for domestic use.

Steve A.
User avatar
Steve Anderson
"Fester! Don't do that to 'Thing'"
 
Posts: 3951
Joined: Fri Mar 30, 2007 10:54 pm
Location: Bangkok, Thailand

Test film.

Postby Stephen » Thu May 17, 2007 11:31 pm

DrZarkov wrote:I've found a picture of the first german TV test movie, a 10 minute silent movie calles "Wochenende" (weekend) with Imogen Orkutt and Schura von Finkelstein. They are singing (in a silent movie!?) "Horch, was kommt von draußen rein". This film had been used from 1929 to around 1934.
I am surprised how good this picture looks in the 30 line (3.0 daN) lateral scan format. This is remarkable for a 1200 pixel image.
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Postby AncientBrit » Thu May 17, 2007 11:42 pm

Stephen,

Is a 3.0 DAN equivalent to a Black Belt??!

(Sorry)

Regards,

GL
AncientBrit
Green padded cells are quite homely.
 
Posts: 858
Joined: Mon Mar 26, 2007 10:15 pm
Location: Billericay, UK

Postby Stephen » Thu May 17, 2007 11:54 pm

AncientBrit wrote:Stephen,

Is a 3.0 DAN equivalent to a Black Belt??!

(Sorry)

Regards,

GL
Actually, I was just "fooling around", referring to my proposal for a new standard unit of measure for television scanning lines, the nipkow (or possibly baird). See the thread at http://www.taswegian.com/NBTV/forum/vie ... 1&start=15 . 30 lines would be 3.0 decanipkow (daN). :wink:
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Postby gary » Fri May 18, 2007 12:51 am

DrZarkov wrote:My wife saw me drawing, and she said that our testcards are looking so "eighties", we should take something more old-fashioned. And she said that a moving picture would make a better testcard. She made one example with Photoshop as a gif-anim. I was not able to transfer it into anything which "AVI to NBTV" will convert until now, maybe somebody knows a program for that. Most video converting programs have a big problem with a video of only 32 x 48 pixels...


Adding animated GIFs to Video2NBTV anon (watch this space).
User avatar
gary
"Fester! Don't do that to 'Thing'"
 
Posts: 2303
Joined: Sat Jan 27, 2007 11:29 am
Location: Bundanoon, Australia

Postby DrZarkov » Fri May 18, 2007 2:50 am

@Stephen: "Baird" is to measure vertical scanning, and "Nipkow" for horizontal scanning. I suggest not simply to count the lines, that would be too easy. We should create a formula to include the aspect ratio and the frame rate. If interlacing is used, it is nor "Baird" neither "Nipkow", but "Farnsworth". O.k.?

So: 1 Baird = 30 x 3 x 7 x 12,5
1 Nipkow = 30 x 3 x 4 x 12,5
1 Farnsworth = 441 x 3 x 4 x 25

:lol:

I wish you much pleasure to calculate a vertical picture with 60 lines and 38,75 frames: how much is that in "Baird"?
User avatar
DrZarkov
I think I've had a cranial implosion.
 
Posts: 1037
Joined: Mon Feb 19, 2007 11:28 pm
Location: Kamp-Lintfort, Germany

Honouring the pioneers.

Postby Stephen » Fri May 18, 2007 3:43 am

The more I think of it, a standard unit for scanning lines should honour Paul Nipkow, since I believe that he was the first to propose pixel-by-pixel, line-by-line sequential image scanning of any kind. One scanning line could be one nipkow (N). John Logie Baird did much experimentation with different aspect ratios, so he should probably be honoured with a standard unit of measure for that. One baird could be the ratio of lateral size to vertical size times ten to avoid small figures. Philo Farnsworth did not have much to do with interlacing. Actually, Mr. Baird was the first to experiment with it. However it was Ulises Sanabria that really promoted his version of interlacing in North America, a triple interlace 45 line scheme. Perhaps then he should be honoured with a standard unit of measure for frames per second (fps). One fps could be one sanabria.

Therefore a 441 line 4:3 aspect ratio 25 fps system becomes:

4.41 hectonipkow (hN), 1.33 decabaird (daB), 2.5 decasanabria (daS)

The standard club format becomes:

3.2 decanipkow (daN), 6.66 baird (B), 1.25 decasanabria (daS)

This makes perfect sense. It is almost intuitive. The NBTVA should quickly adopt this system! :wink:
Stephen
User avatar
Stephen
Anyone have a spare straightjacket?
 
Posts: 427
Joined: Fri Jan 26, 2007 7:00 am

Next

Return to Mechanical NBTV

Who is online

Users browsing this forum: No registered users and 3 guests